3.7 Research method – qualitative versus quantitative
3.7.1 Adoption of mixed methods
For many years the advocates of quantitative and qualitative research
approaches have been involved in keen debate about the superiority of one
method over the other (Johnson and Onwuegbuzie, 2004). Evidently, both
methods have their limitations. However, it is considered that biases inherent
in one method could counteract or annul the biases in the other hence, a means
for seeking convergence across qualitative and quantitative methods became
established (Creswell, 2009). The goal of ‘mixed methods’ is to draw from the
strength of each method and minimise the weaknesses of both in single
research studies and across studies (Johnson and Onwuegbuzie, 2004).
Mixed methods have been widely used in tourism research (see Walle, 1997;
Jenkins, 1999) because it is beneficial to combine different field methods in
order to minimise the weaknesses of individual approaches. Creswell (2009)
suggests the use of multi-methods strategy, as this would improve the
researcher’s ability to determine the accuracy of findings as well as convince
the audience of that accuracy. The position of the post-positivist, particularly a
critical realist, is that all measurements are fallible hence it becomes important
to employ multiple measures and observations and use triangulation to obtain a
better representation of reality.
The term triangulation refers to the use of multiple methods to corroborate the
reliability of a particular research tool and the validity of the data collected
(McNeill and Chapman, 2005). Typically this will involve combining
quantitative and qualitative methods to ensure the accuracy of the data gathered
through each method (McNeill and Chapman, 2005). Cresswell (2009)
suggests triangulating different data sources of information by examining
evidence from the sources and using it to build a coherent justification for
themes. The process where themes are established on the basis of converging
several sources of data or perspectives from participants can add to the validity
of the study (Cresswell, 2009). In this study, the development of the attraction
quality construct was based on this process (see scale development for a
detailed discussion).
The mixed-method approach is not without its demerits. A multi-methods
approach can be expensive and produce enormous amounts of data which can
be difficult to analyse (Devine and Heath, 1999). Devine and Heath (1999)
stress that multiple methods are likely to come up with incongruous findings,
which results in the question of what to do with the data (whether to merge all
or ignore some of the data). However, there are few areas of tourism research
such as holidaying, visitation of attractions and consumption of other leisure
related products like hotel stay and eating-out, where one research method
alone is sufficient, and consequently, social science research is becoming
increasingly pluralistic (McNeill and Chapman, 2005). While people’s actions
are a result of their interpretation of the situation, their interpretations and their
choices may also be limited by structural factors external to them and beyond
their control as in the case of the performance (quality) of visitor attractions
(McNeill and Chapman, 2005).
The use of multiple methods can be introduced to a study at any stage of the
research process – from the initial exploration of the topic or concept(s)
through data collection to the analysis stage (Creswell, 2009). It has been
advised that researchers should carefully consider the reason for choosing
individual methods, the study aim and hypotheses and the main philosophy
underpinning the investigation (Creswell, 2009, Tashakkori and Teddlie, 2003
and Greene and D’Oliveira, 1999). Creswell (2009) identifies four important
aspects that influence the design procedure for a mixed methods study. These
factors are timing, weighting, mixing and theorizing (transforming)
perspectives.
The issue of timing relates to whether the gathering of the two types of data
(qualitative and quantitative) will be in phases or will run concurrently. Where
the data are collected in phase either of the two can come first depending on
the intent of the researcher or the nature of the study. Currall and Towler
(2003) posits that when qualitative data are collected first, the intent is likely to
be that the researcher wants to explore the topic with the aim of expanding the
understanding through a second phase in which data are collected from a large
number of subjects. This idea is in line with the procedure predominantly
followed in this research.
On the other hand, in some cases, it may be undesirable or unpractical to
collect qualitative and quantitative data at different times. In this case, data are
collected concurrently and simultaneously implemented. In the second phase of
the current study, qualitative data were collected concurrently with quantitative
data using open ended questions in the primary research instrument (see further
detail in questionnaire design).
The second aspect identified by Creswell (2009) is weighting, which connotes
the priority. Depending on the nature of the study, the weight might be equal or
one method may feature more dominantly over the other. Creswell (2009)
submits that the priority given to any of the two methods will be determined by
a) the research’s interest; b) the audience of the study and c) what the
investigator intends to emphasise in the study. In this study more emphasis is
on the quantitative data owing to the focus of the study (enumerated in chapter
one)
Mixing is the third aspect and it entails the amalgamation of the two methods at
some point(s) in the research. According to Creswell (2009), mixing will
normally take place in one of three forms. Either the data are combined at one
end of the continuum, kept separate at the two ends, or blended in some way
between the two extremes. It is possible to have the two data bases kept
separate, however, they must be connected. As in this project, the results from
the initial qualitative research were used as the basis of developing the research
instrument for the second phase, hence connecting the first phase to the second.
Connecting here denotes a mixing of both qualitative and quantitative research
whereby the connection takes place between the data analysis of the first
research phase and the data collection of the second research phase. Walle
(1997) pointed out that this approach is widely used in tourism and marketing
studies but noted that rigorous quantitative methods are being supplemented
with qualitative method to address the loss of relevance amongst practitioners.
In this light and as stated earlier, some qualitative data are gathered
concurrently in the second phase along with a large amount of quantitative
data. The qualitative themes were however transformed into counts and
compared with descriptive quantitative data. In this case, the mixing involves
integrating the qualitative and quantitative data by combining the two.
The fourth aspect, theorizing, is concerned with the theoretical perspective that
guides the entire design of the study. Typically, theories used in guiding the
research are contained in the sections detailing framework that shapes the type
of questions asked, who participates in the study, how data are gathered, and
the implications made from the study (Creswell, 2009). The theoretical lens
utilised in this study is the hierarchical perspective to evaluation of service
quality perception, modelling quality as a formative construct (see 2.8.4.4)
using a ‘performance-only’ framework (see 2.12).
A number of typologies and classifications of mixed methods exist. Creswell et
al., (2003) advanced six types of mixed methods strategies:
• Sequential explanatory strategy;
• Sequential exploratory strategy;
• Sequential transformative strategy;
• Concurrent triangulation strategy;
• Concurrent embedded strategy;
• Concurrent transformative strategy.
In this research, a sequential exploratory mixed methods approach was
adopted; involving a first phase of qualitative data gathering and analysis,
followed by a second phase of quantitative data gathering and analysis based
on the results of the initial qualitative data. The sequential exploratory strategy
is particularly advantageous when research is building a new instrument and
confirming a scale (Creswell, 2009; Devellis, 2003 and Churchill, 1979), like
the one built in this study to determine quality attributes of visitor attractions.
In addition to the sequential exploratory procedures, a slight concurrent
embedded mixed methods perspective was incorporated at the second phase.
Concurrent embedded mixed methods procedures are those in which the
researcher collects both quantitative and qualitative data simultaneously in one
phase. In this type of design, the researcher collects both forms of data at the
same time and then integrates the information of the overall results. For this
reason, two sets of questionnaires which included both quantitative and 129
qualitative questions were designed for the current research and used in
concurrent surveys.